随着移动互联网等信息技术的快速发展，数据呈爆炸式增长。“大数据”的兴起对云存储系统的发展和应用起着积极的推动作用，而云存储系统的性能和可用性是影响云存储服务应用和推广的重要因素。重复数据删除技术可以有效地提高存储效率和网络带宽利用率，并已逐渐被应用于云存储系统中。本项目拟通过研究云存储系统中的重复数据删除技术来提高云存储系统的性能和可用性，主要研究内容包括：（1）测试和分析缓存空间分配对重复数据删除存储系统读写性能的影响；（2）在基于重复数据删除技术的云存储系统节点内部研究基于应用负载变化的动态缓存管理策略，提高缓存利用率和存储系统性能；（3）研究在多个云存储系统之间根据重复数据删除技术中的数据引用特征进行数据的布局，提高云存储系统的可用性。通过本申请项目的研究，可以有效地提高云存储系统的性能和可用性，并推动重复数据删除技术在云存储系统中的更广泛应用。

Data are exploding with the rapid development of the information technology, such as the mobile Internet. The rise of "big data" further promotes the wide development and application of cloud storage systems. The performance and availability of cloud storage systems are important factors that affect the applications and the popularizations of the cloud storage services. The data deduplication technology can effectively improve the storage efficiency and reduce the amount of the data transmission. Recently, it has been gradually applied to cloud storage systems. This project intends to research on the data deduplication technology in cloud storage systems to improve the performance and availability. The main research contents include: (1) Experimental evaluation and analysis of the effect of the cache space allocation on the read and write performance of deduplication storage systems. (2) Research on the dynamic cache management strategy based on the application patterns inside cloud storage systems to improve the cache efficiency and the system performance. (3) Research on the data layout strategy among multiple cloud storage systems based on the data reference characteristics to improve the system availability. The goal of the project is to effectively improve the performance and availability of cloud storage systems, and further promote the applications of the data deduplication in cloud storage systems.